The Research On Drowsiness Detection System

The technology of face detection and recognition is a technology that uses computer to analyse images and discriminate identity or recognize status from the input image. With the rapid development of computer technology,many various areas have used this technology,such as ID card,intelligent entrance control in buildings and so on.This paper applies the technology of face recognition in drowsiness detection system and tries to warn the driver when drowsiness is detected,so as to avoid serious traffic accidents.In this paper, we first present the research background of drowsiness detection system ,then we introduce the study actuality of drowsiness detection. Next,we describe our system in detail.In order to improve efficiency,we do some necessary pretreatments before face detection,such as changing RGB images to gray level images, geometric emendation, illumination intensity emendation,removing noises and edge detection.First,we use skin color model to detect face in the original image,the biggest connected area is labelled as face area.Then we use a new combination projection of differential coefficient and integral to locate the two eyes separately. Next we detect the edge of eye image using canny arithmetic operators,and calculate the average height of two eyes to judge if they are closed or not.Finally we use PERCLOS standerd from USA to determine drowsiness,and warn the driver if drowsiness is detected. We simulate the whole system using Matlab tools.